Density Functional Theory Calculations in Stereochemical

Subscriber access provided by DUESSELDORF LIBRARIES

Article

Density Functional Theory Calculations in Stereochemical Determination of Terpecurcumins J-W, Cytotoxic Terpeneconjugated Curcuminoids from Curcuma longa L. Xionghao Lin, Shuai Ji, Xue Qiao, Hongbo Hu, Ni Chen, Yinhui Dong, Yun Huang, Dean Guo, Pengfei Tu, and Min Ye J. Org. Chem., Just Accepted Manuscript • Publication Date (Web): 12 Nov 2013 Downloaded from http://pubs.acs.org on November 19, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Organic Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Organic Chemistry

Density

Functional

Theory

Calculations

in

Stereochemical

Determination of Terpecurcumins JW, Cytotoxic Terpene-conjugated Curcuminoids from Curcuma longa L.

Xionghao Lin,† Shuai Ji,† Xue Qiao,† Hongbo Hu,‡ Ni Chen,‡ Yinhui Dong,‡ Yun Huang,† Dean Guo,† Pengfei Tu,† and Min Ye*,†

†

State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical

Sciences, Peking University, 38 Xueyuan Road, Beijing 100191, China ‡

College of Food Science and Nutritional Engineering, China Agricultural University, 17

Qinghua East Road, Beijing 100083, China

Corresponding author *Tel/Fax: +86 10 82802024. E-mail: [email protected] (M. Ye).

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 44

Table of Contents

15 10

exptl. for 1

calcd. for 1

5 0 -5 -10 -15 190

calcd. for ent-1 240

290

340

390

Wavelength /nm


440

490

Page 3 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT: Fourteen novel terpene-conjugated curcuminoids, terpecurcumins JW (114), have been isolated from the rhizomes of Curcuma longa L. Among them, terpecurcumins J–R and V represent four unprecedented skeletons featuring an unusual core of hydrobenzannulated[6,6]-spiroketal (1 and 2), bicyclo[2.2.2]octene (37), bicyclo[3.1.3]octene (8), and spiro-epoxide (13), respectively. The structures of compounds 114 were elucidated by extensive spectroscopic analysis, and their absolute configurations were established by electronic circular dichroism (ECD), vibrational circular dichroism (VCD) and

13

C NMR spectroscopic data analysis, together with

density functional theory (DFT) calculations. The structure and configuration of 1 was further confirmed by single crystal X-ray diffraction (CuK). The biogenetic pathways of 114

were

proposed,

involving

Michael

addition,

condensation,

Diels-Alder

cycloaddition, and electrophilic substitution reactions. Terpecurcumins showed more potent cytotoxic activities than curcumin and ar-/-turmerone. Among them, terpecurcumin Q (8) exhibited IC50 of 3.9 M against MCF-7 human breast cancer cells, and mitochondria-mediated apoptosis played an important role in the overall growth inhibition. Finally, LC/MS/MS quantitative analysis of five representative terpecurcumins indicated these novel compounds were present in C. longa at ppm level.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

INTRODUCTION The global burden of cancer has been growing each year.1 On the other hand, effective drugs for the prevention and treatment of cancer are still in shortage.2,3 Therefore, the exploration of new oncology drugs with fewer adverse reactions remains to a big challenge for pharmaceutical chemists. Given their unique structures, natural products are a major source for the discovery of oncology drugs.4

Turmeric is derived from the rhizomes of Curcuma longa L. (Zingiberaceae family). It is a popular natural spice and a well known Chinese herbal medicine. The major chemical constituents of turmeric are curcuminoids and terpenes, which exhibit diverse biological activities, including anti-cancer, anti-inflammatory, antioxidant, and antiviral activities.5,6 Among them, the most abundant and intensively studied is curcumin, one of the most well known natural products as a promising anticarcinogenic agent.79 Recently, we isolated a series of minor compounds terpecurcumins AI from C. longa, which represented a series of structures conjugating curcuminoids with sesquiterpenes through CC or COC bond.10 Similar structures were also reported by Huang and coworkers.11,12 Interestingly, some of these new compounds showed higher cytotoxic activities than curcumin against human cancer cells.

In our continued efforts to discover anticancer compounds from C. longa, terpecurcumins JW (114) (Figure 1) were isolated from a less polar fraction of turmeric extract. Among them, compounds 18 and 13 possess four unprecedented skeletons that are structurally distinct from the previously reported terpecurcumins A–I.10 The structures of 114 were identified on the basis of extensive spectroscopic analysis, and their absolute configurations were unambiguously established by single crystal X-ray diffraction (for 1) and quantum chemical calculations including electronic circular dichroism (ECD), vibrational circular dichroism (VCD), and 13C NMR spectra prediction.


Page 4 of 44

Page 5 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Terpecurcumins represent a new class of turmeric compounds, and indicate novel biosynthetic pathways. In this paper, we report the isolation, structural elucidation, plausible biogenetic pathways, quantitative analysis, and cytotoxic activities of terpecurcumins JW.

Figure 1. Structures of terpecurcumins JW (114) isolated from Curcuma longa L.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RESULTS AND DISCUSSION Structural Elucidation and Absolute Configuration of Terpecurcumins JW (1–14). Terpecurcumin J (1) has a molecular formula of C34H40O6 as established by positive HRESIMS ([M+H]+ m/z 545.2906, calcd 545.2897), requiring fifteen degrees of unsaturation. The IR spectrum showed abundant absorption bands for hydroxyl group (3378 cm1), carbonyl group (1734 cm1), and phenyl ring (1606, 1513 and 1446 cm1). The

13

C NMR spectrum of 1 showed eighteen sp2 resonances (twelve of which were

assigned to two A2B2 aromatic rings), a ketone signal (C 206.8), and a characteristic quaternary signal (C 98.0) for a doubly oxygenated carbon (Table 1). The remaining unsaturations in compound 1 should be due to three rings. 1H1H COSY, HSQC and HMBC spectral analysis for 1 indicated the presence of three substructures 1a (from 4-OH to C-1 and C-2), 1b (from C-3 to 8-OH), and 1c (from C-2 to 8-OH) (Figure S1). Substructure 1a was established by starting from an oxygenated cyclohexene ring with a methyl substituent, which was characterized by the HMBC correlations of H3-15/C-2, C-3 and C-4, H-2/C-4 and C-6, and H2-5/C-1 and C-4. The HMBC correlations of H3-14/C-6, C-7, C-8 and C-9, and of H-8/C-9, as well as the 1H1H COSY correlations of H-6/H-7/H-8 and H3-14/H-7 established the linkage of C-6 and C-7, the latter of which was further connected to C-8 and C-9. The methyl signals at

H 1.10 (H3-12) and 1.32 (H3-13) exhibited long-range correlations with C-10, C-11, and C-1. This evidence suggested the quaternary carbon (C-11), which bears two methyl groups, was situated between C-10 and C-1. In addition, the HMBC correlations of H3-12/C-9, H3-13/C-9 and C-2, and H-10/C-9 indicated the connection of C-9 to C-2. The chemical shifts for C-2 (C 146.5) and C-1 (C 73.0) (Table 2), as well as the HMBC correlation of H-8/C-2, suggested that the spiroketal carbon C-9 (C 98.0) was simultaneously linked to C-1 and C-2 through ether bonds, and thus established the substructure 1a containing a hydrobenzannulated[6,6]-spiroketal core. The substructure 1b was established by the 1H1H COSY correlations of H-3/H-4, H-6/H-7, and


Page 6 of 44

Page 7 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


H-9/H-10, as well as the HMBC correlations of H-6 and H-10 with C-8 and C-4, and of H-3 with C-5. Substructures 1a and 1b were connected through C-2C-3, according to the HMBC correlations of H-4 with C-2, and of H-3 with C-1 and C-9. The remaining substructure 1c contained a p-hydroxybenzyl group according to the NMR data (Table 1). The HMBC correlations of H-3 with C-2, of H-4 with C-2 and C-3, and of H-6 and H-10 with C-4 indicated the connection of C-2C-3C-4C-5. Substructures 1a and 1c were believed to connect through C-1C-2 bond, albeit the expected HMBC correlation between H-3 and C-1 was too weak to be observed. Based on the above deductions, the planar structure of 1 was established as shown in Figure 1.

The relative configuration of 1 was determined based on NOE spectra and coupling constants of key proton signals (Figure S2). The NOE enhancements of H-1(H 4.44, d, 2.4 Hz) with H-6 and H-5b, of H-4 with H-5a, H-7 and H3-14, and of H-5a with H3-14 indicated that the A/B ring junction has a cis configuration. Thus, H-1, H-6, H-5b and 4-OH were oriented on the same side of A ring, while H-4, H-5a, H-7 and H3-14 were on the other side. Likewise, the NOE enhancements of H-2/H-7, H-6/H-8b, and H-7/H-8a suggested that H-1, H-6 and H-8b were oriented on the same side of B ring, while H-2, H-7 and H-8a were on the other side. Although the NOE enhancements of H3-12/H-10b, H3-13/H-10a, and H3-13/H-2 were observed, configuration of the spiro junction of ring B and ring C was still unable to be established by NOE experiment due to signal overlapping of H-8a and H-10a.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 8 of 44

Table 1. 1H NMR Data for Terpecurcumins J−P (17) in Acetone-d6 Recorded at 600 MHz (δ in ppm, J in Hz). No. 1 3 4 6 7 9 10 11

1

5

2.36 (m) a 1.52 (m) b 1.71 (m) 2.29 (m) 1.75 (m) a

4.44 (d, 2.4) 5.20 (brs) 4.21 (t, 4.8)

5

6

7

6.69 (d, 1.7)

5.69 (s) 3.02 (d, 7.6) 3.07 (dd, 7.6, 1.7) 6.97 (d, 1.5)

5.85 (s) 2.69 (m) 3.35 (d, 6.0) 6.77 (d, 2.4)

5.72 (s) 2.94 (d, 6.0) 3.13 (d, 6.0) 6.81 (m)

6.71 (m) 6.82 (d, 1.7) 3.71 (s) 6.33 (d, 15.5) 6.77 (m) 6.78 (m)

6.80 (d, 8.2) 6.82 (d, 8.2) 3.86 (s) 6.52 (d, 15.8) 7.47 (d, 15.8) 7.27 (d, 1.5)

6.71 (d, 8.1) 6.64 (dd, 8.1, 1.8) 3.90 (s) 6.55 (d, 15.8) 7.50 (d, 15.8) 7.28 (d, 1.6)

6.80 (d, 8.1) 6.82 (m) 3.86 (s) 6.52 (d, 15.8) 7.47 (d, 15.8) 7.27 (d, 1.3)

6.68 (d, 7.8) 6.62 (dd, 7.8, 2.4) 3.80 (s) 6.56 (d, 16.2) 7.51 (d, 16.2) 7.30 (d, 1.8)

6.81 (m) 6.94 (s) 3.85 (s) 6.51 (d, 16.2) 7.43 (d, 16.2) 7.26 (d, 1.2)

6.80 (m) 7.03 (d, 1.7) 3.82 (s) 4.40 (brs) 5.19 (s) 2.08 (m) a 1.70 (m) b 2.04 (m) a 1.59 (m) b 1.57 (m) 2.36 (m) 1.78 (m) a

6.85 (d, 8.2) 7.11 (dd, 8.2, 1.5) 3.89 (s) 3.06 (d, 6.2) 5.67 (d, 6.2) 2.33 (s)

6.86 (d, 8.2) 7.12 (dd, 8.2, 1.6) 3.81 (s) 2.99 (d, 6.5) 6.06 (d, 6.5) 2.38 (s)

6.85 (d, 8.2) 7.10 (dd, 8.2, 1.5) 3.88 (s) 2.61 (d, 6.5) 5.96 (d, 6.5) 2.75 (s)

6.86 (d, 7.8) 7.13 (dd, 7.8, 1.8) 3.90 (s) 2.68 (m) 5.80 (d, 6.0) 2.68 (m)

6.84 (d, 8.4) 7.09 (dd, 8.4, 1.2) 3.87 (s) 2.99 (d, 6.0) 5.66 (d, 6.0) 2.33 (s)

1.83 (m) a 0.77 (m) b 1.71 (m) 1.13 (m) 1.56 (m) a

1.88 (m) a 0.96 (m) b 1.65 (m) 1.10 (m) 1.43 (m) a

1.31 (m) a 0.99 (m) b 1.67 (m) 1.08 (m) 1.93 (m) a

1.88 (m) a 0.86 (m) b 1.65 (m) 1.10 (m) 1.48 (m) a

1.85 (m) a 0.80 (m) b 1.75 (m) 1.68 (m) 2.66 (dd, 15.0, 4.2) a

7.11 (d, 9.0) 6.77 (d, 9.0) 6.77 (d, 9.0) 7.11 (d, 9.0)

7.06 (d, 9.0) 6.80 (d, 9.0) 6.80 (d, 9.0) 7.06 (d, 9.0)

4 5.88 (s) 2.75 (dd, 6.4, 2.2) 3.38 (d, 6.4) 6.80 (d, 1.8)

3.01 (m, 2H)

6.14 (d, 15.6) 6.72 (d, 15.6)

3 5.73 (s) 2.96 (d, 6.5) 3.13 (d, 6.5) 6.95 (d, 1.8)

3.03 (m, 2H) 2.91 (m, 2H)

3 4 6 7 9 10 11 1 2 4

6 7 8

2

2.90 (m, 2H)


Page 9 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


1.29 (m) b

1.33 (m) b

1.20 (m) b 2.09 (m) a 1.92 (m) b 5.15 (t, 7.1)

1.16 (m) b 1.98 (m) a 1.72 (m) b 5.08 (t, 6.9)

1.04 (m) b 1.72 (m, 2H)

1.75 (m) a 1.58 (m) b 1.10 (s) 1.32 (s) 1.02 (d, 6.6) 1.47 (brs)

1.76 (m) a 1.59 (m) b 1.12 (s) 1.32 (s) 0.98 (d, 6.5) 1.45 (s)

1.67 (s) 12 1.63 (s) 13 0.82 (d, 6.2) 14 1.86 (d, 1.1) 15 a b H was at the lower field. H was at the higher field.

1.62 (s) 1.57 (s) 0.78 (d, 6.5) 1.74 (d, 1.3)

9 10

2.15 (m) b

5.00 (t, 7.0)

1.12 (m) b 2.04 (m) a 1.88 (m) b 5.10 (t, 7.2)

1.61 (s) 1.47 (s) 0.73 (d, 6.4) 1.81 (d, 1.1)

1.64 (s) 1.59 (s) 0.80 (d, 6.6) 1.95 (d, 1.2)

1.86 (s) 2.10 (s) 0.80 (d, 6.0) 1.85 (s)

6.15 (s)

Table 2. 13C NMR Data for Terpecurcumins J−P (17) in Acetone-d6 Recorded at 150 MHz (δ in ppm). No.

1

2

3

4

5

6

7

1 2 3 4 5 6 7 8 9 10 11

125.2 (qC) 206.8 (qC) 50.3 (CH2) 29.6 (CH2) 133.4 (qC) 130.5 (CH) 116.0 (CH) 156.5 (qC) 116.0 (CH) 130.5 (CH)

2

146.5 (qC)

125.3 (qC) 206.8 (qC) 50.3 (CH2) 30.6 (CH2) 133.8 (qC) 121.6 (CH) 148.2 (qC) 145.6 (qC) 115.7 (CH) 112.8 (CH) 56.0 (CH3) 147.4 (qC)

100.4 (CH) 204.5 (qC) 56.2 (CH) 45.7 (CH) 135.4 (qC) 113.0 (CH) 148.3 (qC) 145.9 (qC) 115.6 (CH) 120.8 (CH) 56.3 (CH3) 177.2 (qC)







1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

118.5 (CH) 132.0 (CH) 128.9 (qC) 129.3 (CH) 116.3 (CH) 158.5 (qC) 116.3 (CH) 129.3 (CH) 73.0 (CH) 129.3 (CH) 136.7 (qC) 66.3 (CH) 34.5 (CH2) 42.6 (CH) 23.2 (CH) 44.8 (CH2) 98.0 (qC) 48.6 (CH2) 32.4 (qC) 31.1 (CH3) 29.3 (CH3) 19.3 (CH3) 20.2 (CH3)

119.4 (CH) 132.5 (CH) 129.6 (qC) 121.1 (CH) 148.4 (qC) 147.9 (qC) 116.1 (CH) 111.1 (CH) 56.2 (CH3) 73.1 (CH) 127.6 (CH) 133.1 (qC) 25.9 (CH2) 23.9 (CH2) 40.1 (CH) 22.2 (CH) 44.7 (CH2) 98.1 (qC) 48.9 (CH2) 32.5 (qC) 31.1 (CH3) 29.3 (CH3) 19.3 (CH3) 24.0 (CH3)

120.7 (CH) 140.1 (CH) 128.3 (qC) 111.2 (CH) 148.7 (qC) 149.7 (qC) 116.2 (CH) 123.5 (CH) 56.3 (CH3) 38.3 (CH) 122.5 (CH) 144.9 (qC) 45.0 (CH) 26.5 (CH2) 46.7 (CH) 38.6 (CH) 35.4 (CH2) 25.7 (CH2) 125.9 (CH) 131.3 (qC) 25.9 (CH3) 17.9 (CH3) 17.2 (CH3) 20.0 (CH3)




Page 10 of 44


120.7 (CH) 140.0 (CH) 128.2 (qC) 111.2 (CH) 148.7 (qC) 149.7 (qC) 116.1 (CH) 123.5 (CH) 56.2 (CH3) 38.4 (CH) 122.5 (CH) 145.1 (qC) 44.9 (CH) 26.3 (CH2) 46.5 (CH) 35.7 (CH) 50.3 (CH2) 200.7 (qC) 128.2 (CH) 154.3 (qC) 27.4 (CH3) 20.4 (CH3) 17.9 (CH3) 19.9 (CH3)

Page 11 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In order to fully establish the structure of 1, we obtained its single crystals from a methanol-water (45:55, v/v) solution after repeated attempts. X-ray diffraction experiment (CuK) unambiguously confirmed the structure of 1 and determined its absolute configuration as 1S, 4S, 6R, 7S, 9R (Figure 2).

Figure 2. X-ray crystallographic structure of terpecurcumin J (1).

Terpecurcumin K (2) should have a molecular formula of C36H44O7, according to its HRESIMS spectrum ([M+H]+ m/z 589.3177, calcd 589.3157). Its NMR data were similar to those of 1 except for the presence of two methoxyl signals (H 3.71, 3.82), and one oxygenated methine (C 66.3) in 1 was replaced by a methylene (C 25.9) in 2. The two methoxyl groups were assigned to H3-11 (H 3.71) and H3-11 (H 3.82) according to their HMBC correlations with C-7 (C 148.2) and C-7 (C 148.4), respectively. The methylene signal at C 25.9 was assigned to C-4 by the HMBC correlations of H3-15/C-2, C-3 and C-4, and the 1H1H COSY correlations of H-4/H-5 (Figure S3). Thus, compound 2 was a close analogue of 1. The relative configuration of the hydrobenzannulated [6,6]-spiroketal moiety in 2 was the same as 1. Particularly, the NOE enhancement between CH3-13 and H-2 (Figure S4) indicated that the spiro junction of rings B and C in 2 had the same relative configuration as 1. In addition, compounds 1 and 2 showed very similar ECD spectra (Figure S5), indicating that 2 also had the absolute configuration of 1S, 6R, 7S, and 9R. In order to further confirm the absolute



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 44

configuration, the ECD spectra of 1 and 2 were predicted by time dependent density functional theory (TDDFT) calculations at the B3LYP/6-31G* level in the gas phase and in acetonitrile (MeCN) solution with the IEFPCM model. The absolute configuration of 1 predicted by ECD spectra calculations (Figure 3) gave the same result as X-ray diffraction, suggesting that ECD calculations could be used to establish the absolute configuration of similar hydrobenzannulated[6,6]-spiroketals. As shown in Figure 3, the calculated ECD spectra for 1S, 6R, 7S, 9R-2, in contrast to 1R, 6S, 7R, 9S-2, were similar to the experimental ECD spectrum of 2. Therefore, the absolute configuration of 2 was confirmed to be 1S, 6R, 7S, 9R. 15

20

exptl. for 1

calcd. for 1

15

10

exptl. for 2 calcd. for 2

10 5

5 0

0

-5 -5

-10 -10 -15 190

calcd. for ent-1

calcd. for ent-2

-15 -20

240

290

340

390

440

490

190

240

290

340

390

440

490

Wavelength /nm

Wavelength /nm

TDDFT/B3LYP/6-31G(d)

Figure 3. Comparison of the calculated ECD spectra for (1S, 4S, 6R, 7S, 9R)-1, (1S, 6R, 7S, 9R)-2, and their enantiomers (ent-1 and ent-2), with the experimental ECD spectra for 1 and 2. Experimental spectra for 1 and 2 in MeCN

; Calculated ECD spectra for (1S, 4S, 6R, 7S,

9R)-1 and (1S, 6R, 7S, 9R)-2 at B3LYP/6-31G* level in the gas phase with the IEFPCM model the gas phase

and in MeCN solution

; Calculated ECD spectra for ent-1 and ent-2 at B3LYP/6-31G* level in

and in MeCN solution with the IEFPCM model

.

Terpecurcumin L (3) was obtained as orange amorphous powder. Its molecular formula C36H44O6 was established by HRESIMS ([M+H]+ m/z 573.3224, calcd 573.3211). The NMR spectra suggested 3 should be a hybrid of curcumin (moiety A) and


Page 13 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-zingiberene (moiety B) as illustrated in Figure S6. For moiety A, the trans-olefinic signals for H-3 and H-4 in curcumin were replaced by two vicinal aliphatic methine signals H 3.13 (H-4, d, J=6.5 Hz) and 2.96 (H-3, d, J=6.5 Hz).13 The signals for moiety B were similar to those of ()--zingiberene, a major sesquiterpene previously reported from C. longa.14,15 The cyclohexene ring substituted with a methyl group and an alkene chain [(CH3)2C=CHCH2CH2CH(CH3)] were established by the HMBC correlations of H-1/C-3, CH3-15/C-2, C-3 and C-4, and CH3-12/C-10, C-11 and C-13, as well

as

the

1

H1H

COSY

correlations

of

H-4/H-5/H-6/H-1/H-2

and

H-10/H-9/H-8/H-7/CH3-14. The alkene chain was connected to C-6 of the cyclohexene ring, as evidenced by the HMBC correlations of H-8 and CH3-14 with C-6, and the 1H1H COSY correlation of H-6/H-7 (Figure S6). The connection between moieties A and B through C-3 and C-1 was established by the HMBC correlations of H-3/C-2and C-6, and H-1/C-2, as well as the 1H1H COSY correlation of H-3/H-1. The above signals accounted for 14 of the 15 unsaturation degrees of 3, and indicated a connection between C-4 and C-4. This connection was evidenced by the HMBC correlations of H-4/C-3 and C-5, and H-4/C-3, as well as the 1H1H COSY correlation of H-4/H-4. Thus, terpecurcumin L was deduced to contain an unprecedented skeleton with a bicyclo[2.2.2]octene motif. The relative configuration of compound 3 was established by NOE experiments (Figure S7). Significant NOE enhancements of H-3/H-6, H-10 and H-6, and H-6/H-10 indicated that H-3, H-6 and the benzene ring at C-4 were oriented on the same side of the cyclohexane ring. Likewise, NOE enhancements of H-4/H-15, as well as H-2/H-7 and CH3-14, indicated that H-4, H-2, H-15 and the side chain at C-6 were on the other side. The trans-relationship between H-3 and H-4 complied with the Alder–Stein rule.16 From a biosynthetic point of view, compound 3 could be derived from curcumin and ()--zingiberene via Diels-Alder [4+2] cycloaddition. ()--Zingiberene is a major sesquiterpene in turmeric, and its absolute configuration had been assigned as 6R, 7S by chemical synthesis.17,18



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 44

Taken into consideration the established relative configuration of 3, the ECD spectrum for (3S, 4S, 1R, 4R, 6R, 7S)-3 was calculated by TDDFT, and matched the experimental data very well (Figure 4). Therefore, the absolute configuration of 3 was determined to be 3S, 4S, 1R, 4R, 6R, 7S. 8

12

calcd. for 3

8

exptl. for 3

4

exptl. for 4

calcd. for 4

4

0

0 -4

-4

calcd. for ent-3 190

-8

calcd. for ent-4

-12

-8 290

390

490

590

190

290

390

490

590

Wavelength /nm

Wavelength /nm


Figure 4. Comparison of the calculated ECD spectra for (3S, 4S, 1R, 4R, 6R, 7S)-3, (3R, 4R, 1R, 4R, 6R, 7S)-4 and their enantiomers (ent-3 and ent-4), with the experimental ECD spectra for 3 and 4. Experimental spectra for 3 and 4 in MeCN

; Calculated ECD spectra for (3S, 4S, 1R, 4R,

6R, 7S)-3 and (3R, 4R, 1R, 4R, 6R, 7S)-4 at B3LYP/6-31G* level in MeCN solution with the IEFPCM model

; Calculated ECD spectra for ent-3 and ent-4 at B3LYP/6-31G* level in MeCN

solution with the IEFPCM model

.

Compounds 46 were isomers of 3. Their UV, IR and NMR spectra were very similar to each other, indicating that they all contained curcumin and ()--zingiberene units, albeit the linkage pattern may be different. For terpecurcumin M (4), the HMBC correlations of H-3/C-2 and C-6, H-4/C-3, H-1/C-2, and H-4/C-3, together with the 1H1H COSY correlations of H-3/H-1 and H-4/H-4 suggested that the two units were also linked by C3C1 and C4C4 bonds (Figure S8). The NOE enhancements of H-3/H-2, H-6 and H-10, as well as H-2/H-14 revealed that H-3, H-2, the benzene ring at C-4, and the side chain at C-6 were oriented on the same side of the cyclohexane ring. In the


Page 15 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


meanwhile, the enhancement of H-4/H-6 indicated that H-4 and H-6 were on the other side (Figure S9). Thus, compounds 3 and 4 only differed in the configurations of C-3 and C-4. By comparing the predicted and experimental ECD spectra (Figure 4), the absolute configuration of 4 was determined to be 3R, 4R, 1R, 4R, 6R, 7S. For terpecurcumin N (5), the HMBC correlations of H-3/C-3and C-5, H-4/C-2, H-4/C-2 and C-6, and H-1/C-3 indicated that the curcumin and ()--zingiberene units were linked by C3C4 and C4C1 bonds (Figure S10). This linkage was further confirmed by the 1

H1H COSY correlations of H-3/H-4, and H-4/H-1. The relative configuration of 5

was determined by NOE experiments. The NOE enhancements of H-3/H-6, H-6 and H-10, and of H-6/H-10 suggested that H-3, H-6 and the benzene ring at C-4 were oriented on the same side of the cyclohexane ring, whereas H-4 and H-2 were on the other side (Figure S11). The absolute configuration of 5 was assigned as 3R, 4R, 1R, 4R, 6R, 7S by comparing the calculated and experimental ECD spectra (Figure S12). Terpecurcumin O (6) was a diastereoisomer of 5, according to the HMBC correlations of H-3/C-3 and C-5, and H-4/C-2, as well as the 1H1H COSY correlations of H-3/H-4 and H-4/H-1 (Figure S13). The NOE enhancements between H-3 and H-15, H-3 and H-6, as well as H-3 and H-10 indicated that H-3, H-15 and the benzene ring at C-4 were oriented on the same side of the cyclohexane ring. In accordance, the enhancement between H-4 and H-6 revealed they were oriented on the other side (Figure S14). Thus, compounds 5 and 6 only differed in the configurations of C-3 and C-4. The absolute configuration of 6 was assigned as 3S, 4S, 1R, 4R, 6R, 7S by comparing the predicted and experimental ECD spectra (Figure S12).

Terpecurcumin P (7) is an oxygenated analogue of 3. Its NMR spectra showed a carbonyl signal at δC 200.7, which was assigned to C-9 according to the HMBC correlations of H-8 and H-10/C-9 (Figure S15). In accordance, C-8 (∆δC +15.0) and C-11 (∆δC +23.0) shifted downfield remarkably when compared to 3 (Table 2). The



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

NOE enhancements of H-3/H-6, H-6 and H-10, as well as H-4/H-15 suggested that 7 had the same relative configuration as 3 (Figure S16). The absolute configuration of 7 was determined to be 3S, 4S, 1R, 4R, 6R, 7S by ECD spectra calculation (Figure S12).

Terpecurcumin Q (8) was obtained as orange amorphous powder. Its molecular formula C36H42O6 was established by HRESIMS ([M+H]+ m/z 571.3059, calcd 571.3054), indicating 16 degrees of unsaturation. The NMR spectra (Tables 3 and 4) suggested that 8 should be a hybrid of a curcumin unit (moiety A) and a sesquiterpene unit (moiety B), as illustrated in Figure S17. Moiety B was proven to be an unusual sesquiterpene skeleton with a bicyclo[3.1.3]octene core substituted by two methyl and one isobutenyl groups. DEPT experiment revealed that moiety B consisted of four methyls at δC 26.6, 26.6, 20.0 and 17.8, two methylenes at δC 35.1 and 32.1, seven methines at δC 130.7, 126.3, 41.4, 40.3, 37.4, 36.1 and 35.4, and two quaternary carbons at δC 138.0 and 130.3. The bicyclo[3.1.3]octene core was established by the HMBC correlations of H-1/C-2 and C-3, H-2/C-4 and C-6, H-5/C-7 and C-9, and H-8/C-4, as well as 1H1H COSY correlations of H-1/H-2, and H-4/H-5/H-6/H-7/H-8/H-9 (Figure S17). The isobutenyl group was established by the HMBC correlations of H-10/C-12and C-13, together with CH3-12/C-11 and C-13. The substitution position of CH3-15, CH3-14 and the isobutenyl group was assigned to C-3, C-7 and C-9, respectively. These deductions were supported by the HMBC correlations of CH3-15/C-2, C-3 and C-4, CH3-14/C-7 and C-8, and H-10/C-8 and C-9, as well as 1H1H COSY correlations of CH3-14/H-7, and H-10/H-9. The linkage of moieties A and B was determined to be a CC bond between C-1 and C-1 according to the HMBC correlation of H-1/C-1, as well as the 1H1H COSY correlation of H-1/H-1 (Figure S17). The above deductions established the planar structure of 8. The bicyclo[3.1.3]octene core is formed by C-4/9 linkage of a bisabolane skeleton. This unusual sesquiterpene unit was


Page 16 of 44

Page 17 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


connected to C-1 of curcumin. This was different from the previously reported terpecurcumins AI, where the sesquiterpene unit was substituted at C-9 or 8-OH of curcumin.10

The relative configuration of 8 was determined by NOE experiments (Figure S18). The enhancements of H-1/H-2, H-8b and CH3-14, and H-9/H-4 and H-5b indicated that H-1, H-2, H-8b and CH3-14 possessed the same orientation, while H-4, H-5b, H-6, H-7 and H-9 were in the opposite orientation. The absolute configuration of 8 was determined to be 1R, 4S, 6R, 7S, 9S, by comparing the experimental ECD spectrum with TDDFT calculated ECD spectra (Figure 5). Molecular orbital (MO) analysis of conformer 8-48 (Figure 6) suggested that the positive Cotton effect

at

ca.

206

nm

mainly

originated

from

the

electronic

transitions

MO141LUMO154 and MO144MO155 of the sesquiterpene substructure involving * and * transitions, while the * transitions in the curcumin unit from MO143 and MO147 to LUMO154 may contribute to the weak Cotton effects at 230450 nm. We calculated the ECD spectra of a truncated model of 8 (TM8, Figure 5), and obtained the same absolute configuration for the sesquiterpene unit as described above. Given that the computational workload for the truncated model was much lower than the whole molecule of 8, this strategy will be used to determine the absolute configuration of compound 9.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 18 of 44

Table 3. 1H NMR Data for Terpecurcumins Q−W (814) in Acetone-d6 Recorded at 600 MHz (δ in ppm, J in Hz). No.

8

1 3 4 6 7 9 10 11

4.35 (d, 12.0) 6.94 (m) 7.65 (dd, 16.2, 3.0) 7.33 (brs)

4.43 (s) 6.91 (m) 7.56 (d, 16.0) 7.30 (s)

6.87 (m) 7.21 (m) 3.89 (s) 6.94 (m) 7.65 (dd, 16.2, 3.0) 7.33 (brs)

6.88 (m) 7.19 (d, 8.0) 3.90 (s) 6.91 (m) 7.56 (d, 16.0) 7.30 (s)

6.87 (m) 7.21 (m) 3.89 (s) 3.18 (m) 5.40 (brs) 2.07 (m)

6.88 (m) 7.19 (d, 8.0) 3.90 (s) 5.61 (d, 10.0) 6.11 (d, 10.0) 2.36 (m) a 2.23 (m) b 1.65 (m, 2H)

3 4 6 7 9 10 11 1 2 4 5 6 7 8

1.84 (m) a 1.54 (m) b 1.40 (brs) 1.63 (m) 1.25 (m) a

9

2.20 (m) 1.58 (m) 2.09 (m) a

10 5.99 (s) 6.71 (d, 16.2) 7.62 (d, 16.2) 7.64 (m) 7.00 (d, 8.4) 7.00 (d, 8.4) 7.64 (m) 6.66 (d, 16.2) 7.61 (d, 16.2) 7.56 (d, 8.4) 6.90 (d, 8.4) 6.90 (d, 8.4) 7.56 (d, 8.4) 4.76 (d, 9.0) 5.54 (s) 4.01 (s) 1.80 (m) a 1.56 (m) b 2.22 (m) 1.91 (m) 1.37 (m) a

11 5.97 (s) 6.65 (d, 15.6) 7.58 (d, 15.6) 7.19 (s)

7.02 (s) 3.92 (s) 6.68 (d, 15.6) 7.58 (d, 15.6) 7.30 (s) 6.88 (d, 8.4) 7.16 (d, 8.4) 3.90 (s) 3.85 (d, 9.6) 5.19 (s) 2.07 (m) a 2.00 (m) b 1.73 (m) a 1.42 (m) b 1.72 (m) 1.40 (m) 1.30 (m) a

12 6.01 (s) 6.72 (d, 15.6) 7.63 (d, 15.6) 7.64 (m) 7.02 (d, 8.4) 7.02 (d, 8.4) 7.64 (m)

13

14

5.96 (s) 6.66 (d, 15.6) 7.59 (d, 15.6) 7.25 (s)

5.99 (s) 6.72 (d, 15.6) 7.59 (d, 15.6) 7.31 (s)

6.73 (d, 16.2) 7.62 (m) 7.35 (s)

7.18 (s) 3.89 (s) 6.69 (d, 15.6) 7.58 (d, 15.6) 7.32 (s)

7.15 (m) 7.19 (m) 3.89 (s) 6.73 (d, 15.6) 7.61 (d, 15.6) 7.34 (s)

6.89 (d, 8.4) 7.19 (dd, 8.4, 1.2) 3.93 (s) 4.78 (d, 9.0) 5.53 (s) 4.03 (s)

6.87 (d, 8.4) 7.16 (d, 8.4) 3.91 (s) 3.90 (m) 5.40 (d, 4.8) 2.26 (m, 2H)

1.85 (m) a 1.62 (m) b 2.14 (m) 2.47 (m) 2.47 (m) a

2.16 (m) a 1.44 (m) b

6.89 (d, 7.8) 7.17 (m) 3.93 (s) 4.61 (d, 4.8) 5.78 (d, 4.8) 2.26 (m) a 1.53 (m) b 2.60 (m) a 1.53 (m) b

1.56 (s)

1.36 (s)


Page 19 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


1.03 (m) b 2.51 (m) 4.85 (d, 9.6) 1.63 (s) 1.63 (s) 0.95 (d, 7.2) 1.60 (s)

1.91 (m) b 5.41 (m) 5.84 (d, 15.6) 2.06 (s) 2.06 (s) 0.78 (d, 6.7) 4.71 (d, 8.5, 2H)

1.28 (m) b 2.02 (m, 2H) 5.12 (t, 6.6) 1.65 (m) 1.59 (m) 0.80 (d, 7.2) 1.78 (m)

9 10 12 13 14 15 a H was at the lower field. b H was at the higher field.

1.20 (m) b 1.82 (m, 2H) 4.89 (t, 7.2) 1.52 (s) 1.46 (s) 0.85 (d, 6.6) 1.68 (s)

2.36 (m) b 6.15 (s) 1.85 (s) 2.08 (s) 0.83 (d, 7.2) 1.79 (s)

1.37 (s) 1.83 (s)

1.36 (s) 1.77 (s)

Table 4. 13C NMR Data for Terpecurcumins Q−W (814) in Acetone-d6 Recorded at 150 MHz (δ in ppm). No.

8

9

10

11

12

13

14

1 2 3 4 5 6 7 8 9 10 11

70.2 (CH) 194.7 (qC) 123.7 (CH) 145.0 (CH) 127.4 (qC) 111.8 (CH) 148.7 (qC) 150.6 (qC) 116.2 (CH) 124.6 (CH) 56.3 (CH3) 194.5 (qC) 123.5 (CH) 145.0 (CH)

72.2 (CH) 195.0 (qC) 125.3 (CH) 143.7 (CH) 127.6 (qC) 111.9 (CH) 148.7 (qC) 150.4 (qC) 116.2 (CH) 124.2 (CH) 56.3 (CH3) 195.0 (qC) 125.3 (CH) 143.7 (CH)

101.8 (CH) 184.1 (qC) 122.6 (CH) 140.6 (CH) 128.5 (qC) 130.9 (CH) 116.7 (CH) 161.2 (qC) 116.7 (CH) 130.9 (CH)

101.3 (CH) 184.7 (qC) 121.9 (CH) 141.9 (CH) 127.3 (qC) 108.0 (CH) 148.2 (qC) 147.9 (qC) 133.0 (qC) 123.6 (CH) 56.3 (CH3) 184.1 (qC) 122.2 (CH) 141.1 (CH)

101.8 (CH) 184.1 (qC) 122.7 (CH) 140.6 (CH) 128.6 (qC) 130.9 (CH) 116.8 (CH) 161.0 (qC) 116.8 (CH) 130.9 (CH)

101.5 (CH) 184.6 (qC) 122.1 (CH) 141.8 (CH) 127.0 (qC) 109.5 (CH) 149.4 (qC) 148.6 (qC) 129.6 (qC) 123.8 (CH) 56.4 (CH3) 184.3 (qC) 122.3 (CH) 141.2 (CH)

101.7 (CH) 184.1 (qC) 122.3 (CH) 141.0 (qC) 128.1 (qC) 112.2 (CH) 151.6 (qC) 151.6 (qC) 117.7 (CH) 123.2 (CH) 56.3 (CH3) 183.6 (qC) 122.9 (CH) 141.5 (CH)

2 3 4

184.8 (qC) 122.0 (CH) 141.1 (CH)

184.8 (qC) 122.2 (CH) 141.5 (CH)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a

127.4 (qC) 111.7 (CH) 148.7 (qC) 150.6 (qC) 116.2 (CH) 124.5 (CH) 56.3 (CH3) 35.4 (CH) 126.3 (CH) 138.0 (qC) 40.3 (CH) 32.1 (CH2) 36.1 (CH) 37.4 (CH) 35.1 (CH2) 41.4 (CH) 130.7 (CH) 130.3 (qC) 26.6 (CH3) 17.8 (CH3) 20.0 (CH3) 26.6 (CH3)

127.6 (qC) 111.9 (CH) 148.7 (qC) 150.4 (qC) 116.2 (CH) 124.2 (CH) 56.3 (CH3) 135.7 (CH) 130.3 (CH) 144.5 (qC) 30.8 (CH2) 24.8 (CH2) 40.5 (CH) 38.2 (CH) 38.2 (CH2) 126.8 (CH) 140.4 (CH) 40.5 (qC) 26.8 (CH3) a 26.5 (CH3) a 16.1 (CH3) 110.3 (CH2)

127.6 (qC) 130.9 (CH) 116.7 (CH) 160.5 (qC) 116.7 (CH) 130.9 (CH) 75.9 (CH) 124.5 (CH) 140.1 (qC) 67.5 (CH) 30.8 (CH2) 38.1 (CH) 31.2 (CH) 35.7 (CH2) 26.5 (CH2) 125.5 (CH) 131.5 (qC) 25.8 (CH3) 17.7 (CH3) 15.1 (CH3) 20.8 (CH3)

128.1 (qC) 111.4 (CH) 148.7 (qC) 149.9 (qC) 116.1 (CH) 123.9 (CH) 56.2 (CH3) 38.5 (CH) 126.5 (CH) 134.5 (qC) 31.1 (CH2) 22.9 (CH2) 44.5 (CH) 32.8 (CH) 36.2 (CH2) 26.5 (CH2) 125.4 (CH) 131.2 (qC) 25.7 (CH3) 17.6 (CH3) 15.1 (CH3) 23.6 (CH3)

128.1 (qC) 111.5 (CH) 148.8 (qC) 150.1 (qC) 116.2 (CH) 123.9 (CH) 56.3 (CH3) 75.7 (CH) 124.2 (CH) 140.3 (qC) 67.4 (CH) 31.6 (CH2) 39.2 (CH) 28.9 (CH) 49.7 (CH2) 200.1 (qC) 124.9 (CH) 154.5 (qC) 27.4 (CH3) 20.5 (CH3) 15.4 (CH3) 20.7 (CH3)

Values may be interchanged.


Page 20 of 44

128.1 (CH) 111.5 (CH) 148.7 (qC) 149.9 (qC) 116.2 (CH) 123.7 (CH) 56.3 (CH3) 37.8 (CH) 122.8 (CH) 137.3 (qC) 31.0 (CH2) 24.4 (CH2) 69.2 (qC) 64.3 (qC) 21.3 (CH3) 20.5 (CH3) 23.4 (CH3)

129.5 (qC) 111.5 (CH) 148.8 (qC) 150.0 (qC) 116.2 (CH) 123.8 (CH) 56.3 (CH3) 74.9 (CH) 120.8 (CH) 143.2 (qC) 31.4 (CH2) 23.6 (CH2) 67.1 (qC) 61.8 (qC) 21.2 (CH3) 20.8 (CH3) 23.4 (CH3)

Page 21 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


12

calcd. for 8

40

exptl. for 8

8

20 4 0

calcd. for TM8 exptl. for 8

0

10

calcd. for TM9

5

exptl. for 9

0

-4

-5

-20 -8

calcd. for ent-8

-12 190

calcd. for ent-TM9

calcd. for ent-TM8 -10

-40 290

390

490

590

190

290

Wavelength /nm

390

490

590

190

290


390

490

590

Wavelength /nm

Wavelength /nm


Figure 5. Comparison of the calculated ECD spectra for (1R,4S,6R,7S,9S)-8, the truncated models of (1R,4S,6R,7S,9S)-8 (TM8) and (6R,7S)-9 (TM9), and their enantiomers (ent-8, ent-TM8, and ent-TM9), with the experimental ECD spectra for 8 and 9. Experimental spectra for 8, and 9 in MeCN

; Calculated ECD spectra for (1R,4S,6R,7S,9S)-8, TM8 and TM9 at

B3LYP/6-31G* level in MeCN solution with the IEFPCM model

; Calculated ECD spectra for

ent-8, ent-TM8 and ent-TM9 at B3LYP/6-31G* level in MeCN solution with the IEFPCM model


.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 44

Figure 6. Important molecular orbitals (MO) involved in the key transitions in ECD spectra of conformer 8-48 in MeCN with IEFPCM model at B3LYP/6-31G(d) level.

Terpecurcumin R (9) is an isomer of 8, as deduced from HRESIMS spectrum ([M+H]+ m/z 571.3059, calcd 571.3054). The NMR spectra indicated that compound 9 also possessed a curcumin unit (moiety A) (Tables 3 and 4). The remaining part (moiety B) was deduced to be a sesquiterpene similar to ()--sesquiphellandrene or (+)--turmerone.19 A cyclohexene ring bearing an exocyclic double bond and a side chain (CH3)2CCH=CHCH2CH(CH3) was established by HMBC and

1

H1H COSY

correlations (Figure S19). The HMBC correlations of H-1/C-10, C-12 and C-13, CH3-12/C-1, and CH3-13/C-1 suggested that moieties A and B were connected through C-1 and C-11. Based on the above deductions, the planar structure of 9 was established


Page 23 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


as shown in Figure 1. Given the flexible structure of compound 9, particularly the curcumin moiety, to calculate its ECD spectrum by TDDFT was time-consuming. Based on our experience with compound 8, to calculate the spectrum for the sesquiterpene moiety could obtain the same result, while the computation workload could be remarkably reduced. Thus, ECD spectra of a truncated model of 9 (TM9) were predicted. In 200250 nm region, the calculated ECD spectra for TM9 (6R, 7S) showed similar positive Cotton effects to the experimental spectra for compound 9, and thus determined its absolute configuration as 6R, 7S (Figure 5). Furthermore, we isolated from turmeric a major compound (+)--turmerone (6R, 7S), a close analogue of the sesquiterpene moiety of 9 (Figure S20).20,21 The experimental ECD spectrum of (+)--turmerone was very similar to that of 9, exhibiting a positive Cotton effect at 240 nm (Δε +6.94) (Figure S21). This result also indicated that compound 9 possessed the absolute configuration of 6R, 7S.

Terpecurcumin S (10) was deduced to have a molecular formula of C34H40O5, C2H4O2 less than that of terpecurcumin A.10 The 1H NMR spectrum of 10 displayed no methoxyl signal, and the two aromatic ABX systems in terpecurcumin A were replaced by an A2B2 system in 10 (Table 3), which were further confirmed by HMBC and 1H1H COSY (Figure S22). Thus, compound 10 contained a bisdemethoxycurcumin unit instead of curcumin. The sesquiterpene structure of 10 was the same as that of terpecurcumin A, according to their similar NMR data (Figure S23). Their similar ECD spectra suggested that they had the same absolute configuration (Figure S24). The absolute configuration of terpecurcumin A had been determined in our previous report by ECD spectral calculations.10 Here we used vibrational circular dichroism (VCD) calculation to verify the ECD results. As shown in Figure 7, the VCD spectrum calculated for (1″R,4″S,6″R,7″S) rather than (1″S,4″R,6″S,7″R) showed high similarity with the experimental VCD spectrum for terpecurcumin A, and thus confirmed the absolute



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

configuration of terpecurcumin A as 1″R, 4″S, 6″R, 7″S. The VCD calculation results showed a Confidence Level of 94%, and the ESI value of 45.7 lies at the 35th percentiles in the database for correct assignments (Figure S25).22 The results suggested that both ECD and VCD calculations could predict the absolute configuration of terpecurcumins accurately, in spite of their flexible structures. Based on the above evidences, the absolute configuration of 10 was established as 1″R, 4″S, 6″R, 7″S.

Figure 7. Comparison of the experimental VCD spectrum for terpecurcumin A with calculated VCD spectra for its possible enantiomers (1R, 4S, 6R, 7S) and (1S, 4R, 6S, 7R).

Terpecurcumin T (11) has the molecular formula C36H45O6, as established by HRESIMS. Its NMR spectra were very similar to the previously reported terpecurcumin I,10 except that C-9″ in 11 was a methylene (δC 26.5) instead of a carbonyl group (δC, 200.4) (Tables 3 and 4). This was supported by the triplet of H-10″ (singlet for terpecurcumin I), the upfield shifts of C-8″ (ΔδC 14.8) and C-11″ (ΔδC 23.1), and 1

H−1H COSY correlations of H-8″/H-9″/H-10″ in 11 (Figure S26). The large

trans-diaxial coupling constant (Jaa=9.6 Hz) suggested a trans relationship between H-1″


Page 24 of 44

Page 25 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and H-6″. Considering that compound 11 had very similar planar structure with terpecurcumin I, and that their ECD spectra showed high similarity (Figure S24), the absolute configuration of 11 was assigned as 1″R, 6″R, 7″S.

Terpecurcumin U (12) showed NMR signals for demethoxycurcumin and bisacurone units, and was an isomer of demethoxybisabolocurcumin ether (DMBCE) (Tables 3 and 4).10 The main difference between 12 and DMBCE was location of the methoxyl group (C-7 for 12 and C-7 for DMBCE), which was established by the HMBC correlations of H-1/C-8, and CH3-11/C-7, as well as the 1H1H COSY correlations of H-6/H-7 and H-9/H-10 (Figure S27). The NOE enhancements of H-1″/H-5″b, H-4″/H-5″b, and H-6″/ H-5″a suggested 12 had the same relative configuration as DMBCE (Figure S28). Their similar ECD spectra established the absolute configuration of 12 as 1″R, 4″S, 6″R, 7″S (Figure S24).

Terpecurcumin V (13) has the molecular formula C31H34O7, as established by HRESIMS spectrum ([M+H]+ m/z 519.2374, calcd 519.2377). The IR spectrum indicated the existence of hydroxyl group (3422 cm1), carbonyl group (1756 cm1), phenyl ring (1626, 1599, 1509 and 1454 cm1), and oxirane ring (964 and 843 cm1).23 The NMR spectra (Tables 3 and 4) suggested that 13 should be a hybrid of a curcumin unit (moiety A) and a terpinolene oxide unit (moiety B, a monoterpene), as illustrated in Figure S29.24,25 Moiety B displayed a methine signal at H 3.90 (m, 1H, H-1), an olefinic signal at H 5.40 (d, 1H, H-2), two groups of methylene signals [H 2.26 (m, 2H, H-4); 2.16 and 1.44 (m, 1H each, H-5)], and three methyl signals [H 1.56 (s, 3H, H-8), 1.37 (s, 3H, H-9), and 1.83 (s, 3H, H-10)]. Their connections were established by the HMBC correlations of H-1/C-3and C-7, H-5/C-6 and C-7, CH3-8/C-6, C-7 and C-9, and CH3-10/C-2and C-4, as well as 1H1H COSY correlations of H-1/H-2 and H-4/H-5. Moieties A and B should be linked through a CC bond between C-9 and



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 44

C-1, as evidenced by the HMBC correlations of H-1/C-9, H-1/C-10, and H-10/C-1. Although the relative configuration of 13 could not be determined by NOE experiment due to the peculiar spiro-epoxide structure, the two methyl signals (H 1.56 and 1.37) corresponding to the gem-dimethyl groups at the oxide ring could be distinguished by NOE enhancements of H-1/H-8 and H-5b/H-9 (Figure S30). The above deductions established the planar structure of 13, which has an unprecedented skeleton conjugating curcumin with a terpinolene oxide unit through a CC bond.

The absolute configuration of compound 13 was determined by compulational calculations of ECD spectra and 13C NMR spectra. Bearing two chiral carbons (C-1 and C-6), compound 13 could have four possible stereoisomers [13a (1R, 6R), 13b (1S, 6S), 13c (1R, 6S) and 13d (1S, 6R)]. The calculated ECD spectra for 13a and 13c, rather than 13b or 13d, showed high similarity with the experimental spectrum after a UV correction of 28 nm (Figure 8). To determine the absolute configuration of C-6, we computed

the

13

C

NMR

spectra

for

13a

and

13c

at

the

B3LYP/6-311+G(2d,p)//HF/6-31G(d) level.26 The calculation was particularly focused on the carbon atoms of terpinolene oxide unit because this is the region likely to be influenced by the stereochemistry of the epoxy ring. Moreover, we performed the data evaluation by considering the correlation factors R2 (derived from a linear regression analysis of the calculated vs experimental

13

C chemical shifts), the absolute errors

(|Δδ|=|δexp−δcalcd|), and the mean absolute error (MAE=Σ[|δexp−δcalcd|]/n), which have been successfully used in the relative configuration determination of natural products.27,28 According to Table 5, the higher R2 (0.9982) and lower MAE (1.4 ppm) suggested 13a as the most possible structure for 13. Moreover, the |Δδ| values demonstrated that 13a matched the experimental data better for all carbon signals than 13c. Based on the above evidences, the absolute configuration of 13 was finally established as 13a (1R, 6R).


Page 27 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10

4

calcd. for 13a calcd. for 13c exptl. for 13

5

calcd. for 14c calcd. for 14a

exptl. for 14

2

0

0

-5

-2

calcd. for 13d calcd. for 13b

calcd. for 14b

-10

calcd. for 14d

-4

190

290

390

490

590

200

300

400

500

600

Wavelength /nm

Wavelength /nm


Figure 8. Comparison of the experimental ECD spectra for 13 and 14 in MeCN with those calculated for their possible stereoisomers after a UV correction of 28 nm.

Table 5. Linear Regression Analysis (R2), Absolute Error (|∆δ|), and Mean Absolute Error (MAE) for the Calculated 13C Chemical Shifts of 13a, 13c, 14a and 14c versus the Experimental Values of 13 and 14. No. 1 2 3 4 5 6 7 8 9 10 R2 MAE

δcalcd. (13a)

δcalcd. (13c)

δexp. 13

|∆δ|calcd. (13a)

|∆δ|calcd. (13c)

δcalcd. (14a)

δcalcd. (14c)

δexp. 14

|∆δ|calcd. (14a)

|∆δ|calcd. (14c)

39.7 122.6 139.3 32.6 25.9 66.1 61.7 21.1 19.7 23.3 0.9982

43.2 124.0 138.5 28.8 28.1 64.8 61.4 19.7 20.0 23.6 0.9952

37.8 122.8 137.3 31 24.4 69.2 64.3 21.3 20.5 23.4

1.9 0.2 2.0 1.6 1.5 3.1 2.6 0.2 0.8 0.1

5.4 1.2 1.2 2.2 3.7 4.4 2.9 1.6 0.5 0.2

74.6 120.0 144.2 32.6 24.9 66.7 61.2 20.9 19.4 23.7 0.9996

71.8 116.3 148.6 30.8 25.1 64.9 62.8 20.5 22.0 25.1 0.9959

74.9 120.8 143.2 31.4 23.6 67.1 61.8 21.2 20.8 23.4

0.3 0.8 1.0 1.2 1.3 0.4 0.6 0.3 1.4 0.3

3.1 4.5 5.4 0.6 1.5 2.2 1.0 0.7 1.2 1.7

1.4

2.3

0.8

2.2

Terpecurcumin W (14) is an isomer of 13. The NMR spectral data indicated that 14 also possessed the curcumin and terpinolene oxide units (Tables 3 and 4). The only difference was the linkage pattern. The HMBC correlation of H-1/C-8 in compound 14 suggested these two units were connected through C-8OC-1 rather than C-9C-1



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 44

bond. This deduction was confirmed by the upfield shift of C-9 (∆δ 11.9) and the downfield shift of C-1 (∆δ +37.1). In accordance, H-9 resonated at H 7.15, which showed HMBC correlations with C-5 (δC 129.5) and C-7 (δC 151.6) (Figure S31). The NOE enhancement of H-1/H-9 suggested a cis-configuration between H-1 and C-8 (Figure S32). Similar to compound 13, the absolute configuration of 14 was determined by ECD and

13

C NMR spectra calculations. As shown in Figure 8, the calculated ECD

spectra for 14a (1S, 6S) and 14c (1S, 6R) were consistent with the experimental spectrum for 14, and thus determined the absolute configuration of C-1 as S. The

13

C

NMR spectra for 14a and 14c were then predicted. As shown in Table 5, 14a showed higher R2 value (0.9996 vs 0.9959) and lower MAE value (0.8 vs 2.2) than 14c. Therefore, the absolute configuration of 14 was established as 1S, 6R.

Plausible Biosynthetic Origin of Terpecurcumins JW (114). Terpecurcumins J–W (1–14) constituted a novel class of natural products conjugating the diarylheptanoid skeleton with a sesqui- or monoterpene. Among them, terpecurcumins J–R (1–8) and V (13) represented four unprecedented skeletons featuring an unusual core of hydrobenzannulated[6,6]-spiroketal

(1

and

2),

bicyclo[2.2.2]octane

(37),

bicyclo[3.1.3]octane (8), and spiro-epoxide (13), respectively. Their novel structures prompted us to speculate their biosynthetic pathways, particularly correlations between the new skeletons with known turmeric compounds.

The hydrobenzannulated[6,6]-spiroketal core is very rare in natural products. Terpecurcumins J (1) and K (2) are the first hydrobenzannulated[6,6]-spiroketals isolated from higher plants. More importantly, the novel carbon skeleton is derived from hybridization of a diarylheptanoid and a sesquiterpene unit, and represents a distinct biosynthetic pathway from the recently reported sponge sesterterpenes containing a similar core.2931 Therefore, terpecurcumins J (1) and K (2) have significant novelty not


Page 29 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


only in carbon skeleton, but also in biogenetic origin. They could be derived from -turmerone and dihydrobisdemethoxycurcumin (DHBDMC)/dihydrocurcumin (DHC) through oxidation, ring-opening, Michael addition, and condensation reactions, as depicted in Scheme 1. Initially, -turmerone could be oxidized, and followed by opening of the epoxy ring to produce A. Subsequent Michael addition may hybridize A with DHBDMC and DHC to produce A1 and A2, respectively, which further undergo a key condensation reaction to yield terpecurcumin J (1) and A3. Finally, terpecurcumin J (1) and A3 could be reduced to generate A4 and terpecurcumin K (2), respectively. This speculation was supported by LC/MS/MS analysis of turmeric. The key biosynthetic intermediates A3 and A4 could be detected, though in trace amounts, in the rhizomes of C. longa (Figure S33).

Scheme 1. Plausible biosynthetic pathway for terpecurcumins J (1) and K (2).

Terpecurcumins LP (3–7) represent an unprecedented skeleton containing a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

bicyclo[2.2.2]octene motif resulted from the conjugation of curcumin and a sesquiterpene units via Diels-Alder [4+2] cycloaddition (Scheme 2). Interestingly, compounds 3–6 represent all the four possible stereoisomers. Their structures only differ in the linkage pattern of the two units. For compounds 3 and 4, the two units were linked through C-3C-1 and C-4C-4 bonds. The same units were linked through C-3C-4 and C-4C-1 bonds for compounds 5 and 6. Diels-Alder addition have been extensively reported in natural products.32,33 Recently, Liu et al isolated an enzyme which could catalyze Diels-Alder cycloaddition from a microbial strain.34 The enzymes responsible for the nonstereo-selective biosynthesis of 3–7 in Curcuma longa warrants future study.

Scheme 2. Plausible biosynthetic pathway for terpecurcumins LP (37).

Terpecurcumins Q (8) and R (9) feature a sesquiterpene unit connected to C-1 of curcumin. Particularly, terpecurcumin Q (8) contains an unprecedented carbon skeleton bearing an unusual bicyclo[3.1.3]octene core formed by C-4/9 linking of a bisabolane skeleton. The biosynthetic precursor of 8 could be traced back to ()--zingiberene, which was abundant in the volatile oil of turmeric (Figure S34). ()--Zingiberene could be oxidized, rearranged, and then cyclized (undergoing dehydration) to produce a carbocation, which could then hybridize with the curcumin anion to produce 8. Likewise, compound 9 could be derived from another abundant compound in turmeric oil,


Page 30 of 44

Page 31 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


()--sesquiphellandrene (Figure S35).

Terpecurcumins SU (1012) should have similar biosynthetic pathways with the previously reported terpecurcumins AI.10 The bisabolane units could be derived from

-turmerone/zingiberene, which undertake oxidative rearrangements, and then conjugate to curcuminoid anion by electrophilic substitution (Figure S36). Terpecurcumins V (13) and W (14) represent the first examples of monoterpene-conjugated curcuminoids. They could also be biosynthesized by electrophilic substitution of curcumin with an -terpineol derived carbocation (Figure S37).25

Terpecurcumins JW (114) are curcuminoids hybridized with sesqui- or monoterpenes through different linkage patterns. Given their fairly low amounts in Curcuma longa, whether these compounds are genuine natural products needs to be proved. For this end, turmeric powder was directly analyzed by MALDI-TOF-MS without sample extraction. Majority of the terpecurcumins could be detected, although the pseudomolecular ion peaks were very low in the mass spectra (Figure S38). Furthermore, we macerated turmeric powder in 95% ethanol at room temperature for 5 min, and then analyzed the sample by LC/MS rapidly. A number of terpecurcumins could be clearly detected, and were identified by comparing with reference compounds (Figure S39). These experiments proved that terpecurcumins JW were true secondary metabolites of Curcuma longa, instead of artifacts formed during extraction and separation.

Cytotoxic Activity Evaluation of Terpecurcumins. The overall growth inhibitory effects of terpecurcumins J–W (1–14) were evaluated against human breast cancer cells (MDA-MB-231 and MCF-7), human liver cancer cells (HepG2) and human lung cancer cells (A549) using crystal violet staining method as previously



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

described.35 Paclitaxel was used as the positive control, which showed IC50 values of 0.0390.062 M. Compounds 8 and 9 showed significant cytotoxicities with IC 50 values of 3.912.3 M for all the four cell lines (Table S1). The other compounds were inactive (IC50, 11.860.5 M), while most terpecurcumins showed more potent cytotoxicities than curcumin (IC 50, 31.749.5 M) and ar-/ -turmerone (100 M). It was particularly noteworthy that compound 8 showed IC50 of 3.9 M against MCF-7 cells. Morphologic observations of terpecurcumins-treated MCF-7 cells suggested that induction of apoptosis contributed to the overall growth inhibition. PI single staining demonstrated that compounds 3, 7 and 8 did not perturb cell cycle, but increased the sub-G1 peaks corresponding to apoptotic cells (Figure S40). Exposure of 5 M and 7.5 M terpecurcumin Q (8) for 48 h increased the percentage of sub-G1 cells from 2.03% of control to 26.35% and 28.55%, respectively, which were remarkably higher than curcumin (2.95% for 5 M and 1.86% for 7.5 M). Accordingly, cleaved PARP (polyADP-ribosepolymerase) bands, as a result of caspase digestion, could be clearly observed in western blotting analysis of compounds 3, 7 and 8 treated MCF-7 cells. This result suggested that caspase activation may play an important role in the apoptosis induction of terpecurcumins. To elucidate whether the caspase activation is mediated by the mitochondrial pathway, mitochondria-associated proteins were analyzed by western blotting. Although terpecurcumin Q (8) at 5 M showed little effect on the pro-apototic protein Bax, it could induce the expression of Bak and significantly suppress the anti-apototic proteins Bcl-2 and Bcl-xl. This phenomenon was not observed in the same concentration of curcumin. Taken together, terpecurcumin Q showed potential for the treatment of breast cancer, albeit the mechanism warrants further evaluation.

Quantitative Analysis of Representative Terpecurcumins. Terpecurcumins J (1) and P (7) could be the minor unknown curcuminoids with molecular weight of 544


Page 32 of 44

Page 33 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and 586 Da discovered by Ourisson and coworkers in 1980. However, these compounds were not isolated and identified from Curcuma longa at that time due to low

abundance.36

To

further

understand

the

abundance

of

these

novel

terpene-conjugated curcuminoids in turmeric, a rapid and sensitive LC/MS/MS method was established to simultaneously determine five terpecurcumins 1, 3, 7, 8 and 13, which represented different carbon skeletons (Figure S41). The mass spectrometer was monitored in the selected reaction monitoring (SRM) mode, and all the five analytes showed good linearity (r > 0.9914) (Table S2). The terpecurcumins were detected in all the 14 batches of turmeric at ppm level (generally below 40 g/g, Table S3). Among them, compounds 1, 3 and 8 showed very low abundance with an average of 3.50, 4.45 and 2.41 ppm, respectively. The contents may be closely related to the abundance of their biosynthetic precursors. Due to the low abundance of bisdihydrodemethoxycurcumin in turmeric, compound 1 had a small amount ranging from 0.42 to 6.68 ppm. In addition, contents of the five terpecurcumins varied significantly among different batches. For instance, the content of terpecurcumin V (13) was 146.96 and 134.72 ppm in batches 7 and 13, respectively, and was only 2.73 ppm in batch 8. It is also noteworthy that samples collected from Sichuan Province, which is a long-history major cultivation area for Curcuma longa in China and is generally considered to produce high quality turmeric (also called authentic crude drugs in China), did not contain significantly higher amounts of terpecurcumins than the other samples.

EXPERIMENTAL SECTION Plant Material, Extraction, and Isolation. The one-year-old rhizomes of Curcuma longa L. were collected from Pengzhou City, Sichuan Province, China in December 2009. Identification of the plant was performed by one of the authors (M. Ye). A voucher specimen (No. JH200912) was deposited at School of Pharmaceutical Sciences, Peking



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

University, Beijing, China. Powder of the air-dried rhizomes (30 kg) was extracted with 95% aq. EtOH at 80 C for two times (1 h each time). The extract was concentrated under reduced pressure to afford a residue, which was purified over MCI gel eluted with 80% MeOH to remove the major curcuminoids including curcumin, demethoxycurcumin, and bisdemethoxycurcumin. The sample was then eluted with MeOH to obtain a concentrated extract (291 g), which was separated by silica gel column chromatography (CHCl3/MeOH, 100:0 to 2:1, v/v) to obtain nine fractions (AI) based on TLC analysis. Fraction E (20 g) was chromatographed over silica gel (1.0 kg, 200300 mesh) and eluted with petroleum ether/EtOAc (1:0 to 2:1) to give seven fractions (EAEG). Fraction EC (5.18 g) was purified on an ODS C18 column eluted with MeOH/H2O (50 to 100%, v/v) to obtain six subfractions (ECAECF). ECC was further purified over Sephadex LH-20 and semi-preparative HPLC (YMC Pack ODS-A, 5 m, 250  10 mm i.d., MeCN:H2O=70:30, v/v) to yield terpecurcumins J (1, 9.0 mg) and K (2, 6.0 mg). Likewise, subfractions ECD and ECE were purified by Sephadex LH-20 column chromatography and semi-preparative HPLC (MeCN:H2O=85:15, v/v) to obtain terpecurcumins L (3, 23.3 mg), M (4, 5.1 mg), N (5, 11.3 mg), O (6, 7.2 mg), P (7, 29.3 mg), S (10, 11.0 mg), and T (11, 20.0 mg). Fraction ED (3.74 g) was separated on an ODS C18 column eluted with MeOH/H2O (50 to 90%, v/v) to obtain three subfractions (EDAEDB). EDB was chromatographed over Sephadex LH-20 and then purified by semi-preparative HPLC (MeCN:H2O=70:30, v/v) to obtain terpecurcumins Q (8, 13.0 mg) and R (9, 5.0 mg). Fraction EF (3.15 g) was purified on an ODS C18 column eluted with MeOH/H2O (50 to 90%, v/v) to yield three subfractions (EFAEFC). EFB (965.8 mg) was subjected to Sephadex LH-20 column chromatography eluted with MeOH to afford eight fractions (EFBAEFBH). EFBC was purified by semi-preparative HPLC with MeCNH2O (75:25, v/v) as the mobile phase to yield terpecurcumins U (12, 17.0 mg), V (13, 17.0 mg), and W (14, 3.6 mg). HRESIMS analysis of the pure compounds was conducted on a FT-MS instrument.


Page 34 of 44

Page 35 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Terpecurcumin J (1): pale yellow crystals, mp 202204 C; []23D +2.0 (c 0.08, MeCN); UV (MeCN) max (log ε)=283 (3.95), 324 (4.11) nm; ECD (MeCN) max (∆ε) = 190 (+9.78), 224 (+8.97), 400 (+2.32) nm; IR max=3378, 2954, 2927, 1734, 1707, 1677, 1606, 1513, 1446, 1374, 1263, 1245, 1204, 1044, 825 cm-1; for 1H and 13C NMR data, see Tables 1 and 2; HRESIMS m/z 545.2906 [M+H]+, calcd for C34H41O6, 545.2897.

Terpecurcumin K (2): pale yellow amorphous powder; []23D +1.25 (c 0.08, MeCN); UV (MeCN) max (log ε)=282 (3.99), 327 (4.13) nm; ECD (MeCN) max (∆ε) = 190 (+15.38), 236 (+10.67), 402 (+2.67) nm; IR max=3421, 2956, 1769, 1598, 1514, 1450, 1373, 1242, 1034, 812 cm-1; for 1H and

13

C NMR data, see Tables 1 and 2; HRESIMS

m/z 589.3177 [M+H]+, calcd for C36H45O7, 589.3159.

Terpecurcumin L (3): orange amorphous powder; []21D +130 (c 0.08, MeCN); UV (MeCN) max (log ε)=367 (4.25) nm; ECD (MeCN) max (∆ε) = 197 (6.28), 212 (+2.29), 405 (+2.72) nm; IR max=3419, 2958, 2929, 1769, 1734, 1634, 1577, 1513, 1429, 1374, 1243, 1124, 1034, 969, 813 cm-1; for 1H and

13

C NMR data, see Tables 1 and 2;

HRESIMS m/z 573.3224 [M+H]+, calcd for C36H45O6, 573.3211.

Terpecurcumin M (4): orange amorphous powder; []22D 20 (c 0.04, MeCN); UV (MeCN) max (log ε)=367 (4.20) nm; ECD (MeCN) max (∆ε) = 190 (+17.32), 218 (9.45) nm; IR max=3422, 2959, 2924, 1769, 1734, 1634, 1574, 1512, 1430, 1374, 1242, 1123, 1034, 965, 813 cm-1; for 1H and

13

C NMR data, see Tables 1 and 2; HRESIMS m/z

573.3219 [M+H]+, calcd for C36H45O6, 573.3211.

Terpecurcumin N (5): orange amorphous powder; []22D 27 (c 0.02, MeCN); UV (MeCN) max (log ε)= 368 (4.17) nm; ECD (MeCN) max (∆ε) = 201 (+11.19), 231 (+0.27)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

nm; IR max=3432, 2961, 2926, 1769, 1735, 1633, 1580, 1513, 1429, 1374, 1243, 1126, 1035, 968, 816 cm-1; for 1H and

13


573.3228 [M+H]+, calcd for C36H45O6, 573.3211.

Terpecurcumin O (6): orange amorphous powder; []23D +60 (c 0.04, MeCN); UV (MeCN) max (log ε)= 366 (3.40) nm; ECD (MeCN) max (∆ε) = 190 (9.86), 212 (+7.60) nm; IR max=3410, 2961, 2925, 1734, 1632, 1585, 1513, 1430, 1374, 1262, 1243, 1130, 1037, 968, 815 cm-1; for 1H and

13


573.3226 [M+H]+, calcd for C36H45O6, 573.3211.

Terpecurcumin P (7): orange amorphous powder; []23D +183 (c 0.10, MeCN); UV (MeCN) max (log ε)=365 (4.32) nm; ECD (MeCN) max (∆ε) = 200 (10.69), 212 (+5.43), 247 (+2.32), 365 (+4.43) nm; IR max=3425, 2933, 1678, 1633, 1588, 1513, 1429, 1377, 1270, 1210, 1125, 1033, 970, 816, 756 cm-1; for 1H and 13C NMR data, see Tables 1 and 2; HRESIMS m/z 587.3007 [M+H]+, calcd for C36H43O7, 587.3003.

Terpecurcumin Q (8): orange amorphous powder; []25D +99 (c 0.02, MeCN); UV (MeCN) max (log ε)=344 (4.26) nm; ECD (MeCN) max (∆ε) = 206 (+8.16) nm; IR (KBr)

max=3395, 2969, 2914, 1658, 1579, 1512, 1450, 1429, 1378, 1272, 1161, 1125, 1032, 982, 756 cm-1; for 1H and

13

C NMR data, see Tables 3 and 4; HRESIMS m/z 571.3059

[M+H]+, calcd for C36H43O6, 571.3054.

Terpecurcumin R (9): orange amorphous powder; []25D +166 (c 0.03, MeCN); UV (MeCN) max (log ε)=233 (4.39), 349 (4.35) nm; ECD (MeCN) max (∆ε) = 227 (+5.39) nm; IR (KBr)max=3416, 2957, 2925, 1655, 1579, 1512, 1455, 1429, 1379, 1274, 1163, 1124, 1074, 1031, 980, 771 cm-1; for 1H and

13




Page 36 of 44

Page 37 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Terpecurcumin S (10): orange amorphous powder; [α]D25 +142 (c 0.02, MeCN); UV (MeCN) λmax (log ε)=413 (4.28) nm; ECD (MeCN) max (∆ε) = 200 (+2.98), 403 (+0.34) nm; IR (KBr) νmax=3364, 2961, 2918, 2849, 1735, 1626, 1599, 1509, 1443, 1247, 1169, 1137, 978, 831 cm−1; for 1H and 13C NMR data, see Tables 3 and 4; HRESIMS m/z 529.2943 [M+H]+, calcd for C34H41O5, 529.2948.

Terpecurcumin T (11): orange amorphous powder; [α]D25 +200 (c 0.02, MeCN); UV (MeCN) λmax (log ε)=420 (4.43) nm; ECD (MeCN) max (∆ε) = 202 (+4.15), 405 (+0.79) nm; IR (KBr) νmax=3523, 3427, 2961, 2922, 2854, 1734, 1624, 1592, 1512, 1427, 1294, 1267, 1134, 968, 845 cm−1; for 1H and

13



Terpecurcumin U (12): orange amorphous powder; [α]D25 +170 (c 0.02, MeCN); UV (MeCN) λmax (log ε)=415 (4.37) nm; ECD (MeCN) max (∆ε) = 200 (+3.87), 406 (+0.33) nm; IR (KBr) νmax=3409, 2962, 2935, 1734, 1623, 1597, 1509, 1445, 1246, 1134, 978, 828 cm−1; for 1H and

13

C NMR data, see Tables 3 and 4; HRESIMS m/z 573.2851

[M+H]+, calcd for C35H41O7, 573.2846.

Terpecurcumin V (13): orange amorphous powder; []25D +20.0 (c 0.01, MeOH); UV (MeOH) max (log ε)=202 (4.65), 260 (4.23), 422 (4.76) nm; ECD (MeCN) max (∆ε) = 226 (+5.27), 419 (+1.78) nm; IR (KBr) max=3422, 2923, 1756, 1627, 1599, 1509, 1455, 1279, 1116, 964, 843, 611 cm-1; for 1H and 13C NMR data, see Tables 3 and 4; HRESIMS m/z 519.2374 [M+H]+, calcd for C31H35O7, 519.2377.

Terpecurcumin W (14): orange amorphous powder; []25D +10.2 (c 0.02, MeOH); UV (MeOH) max (log ε)=202 (4.77), 259 (4.28), 419 (4.09) nm; ECD (MeCN) max (∆ε)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

= 200 (+4.09), 422 (+1.70) nm; IR (KBr) max=3424, 2888, 1758, 1629, 1465, 1344, 1279, 1189, 1112, 963, 843, 529 cm-1; for 1H and 13C NMR data, see Tables 3 and 4; HRESIMS m/z 519.2383 [M+H]+, calcd for C31H35O7, 519.2377.

X-ray Crystal Data for Terpecurcumin J (1). C68H79O12 (M =1088.31), monoclinic, space group P21 (no. 4), a = 9.3544(2) Å, b = 28.8050(7) Å, c = 11.3634(3) Å, β = 92.382(2)°, V = 3059.27(13) Å3, Z = 2, T = 180.00(10) K, μ(Cu Kα) = 0.642 mm-1, Dcalc = 1.181 g/mm3, 11980 reflections measured (6.14 ≤ 2Θ ≤ 144.2), 8519 unique (Rint = 0.0295) which were used in all calculations. The final R1 was 0.0581 (>2 sigma(I)) and wR2 was 0.1663 (all data). The goodness of fit on F2 was 1.037. Flack parameter = 0.09 (19). Using Olex2, the crystal structure was solved with the XS structure solution program using Direct Methods and refined with the XL refinement package using Least Squares minimization. The CIF file of X-ray data for 1 has been deposited in the Cambridge Crystallographic Data Centre (deposition no.: CCDC 934684).

ECD Calculations. A preliminary conformational search was performed in SYBYL-X 1.1 using random search method with MMFF94 force field.37 Conformers within 6 kcal/mol were saved and further optimized using the density functional theory (DFT) method at the B3LYP/6-31G* level. Frequency was calculated at the same level of theory. The stable conformers with populations greater than 1% and without imaginary frequencies were submitted to ECD calculation by the TDDFT [B3LYP/6-31G*] method. The excitation energies (E), oscillator strength (f), rotatory strength in velocity form (Rvel), and rotatory strength in length form (Rlen) of the lowest 30 excited states were calculated. Considering solvent effects on calculation, we used the IEFPCM model in MeCN. ECD spectra of different conformers were simulated using SpecDis38 with a half bandwidth of 0.3 eV for compounds 1 and 2 and of 0.4 eV for other compounds. The final ECD spectra


Page 38 of 44

Page 39 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were generated according to the Boltzmann-calculated distribution of each conformer. The calculated ECD spectra were compared with the experimental data. All calculations were performed with Gaussian 09 program package.39

VCD Calculations. A conformational search was carried out with ComputeVOA40 at the molecular mechanics level. Geometry, frequency, IR, and VCD intensity calculations were carried out at the DFT/B3LYP/6-31G* level with Gaussian 09. The calculated frequencies were scaled by 0.97, and the IR and VCD intensities were converted to Lorentzian bands with 6-cm-1 half width for comparison to experiment. In order to accurately compare calculated and experimental IR and VCD data, the recently developed confidence level algorithm CompareVOA22 was used. It allows a direct quantitative comparison of experimental and calculated spectra to provide a measure of agreement. It also provides the enantiomeric similarity index (ESI) for a given configuration, and the position of this value within an internal database for correct VCD assignments.

13

C NMR Chemical Shifts Calculations. The conformers consistent with the NOE

experimental data for diastereoisomers 13a and 13c were optimized by using the Hartree-Fock (HF) method at 6-31G(d) level. The stable conformers with populations greater than or equal to 1% and without imaginary frequencies were used as inputs for the 13

C NMR spectra calculations performed in Me2CO employing the B3LYP function

combined with the 6-311+G(2d, p) basis set. The calculated values of 13C chemical shifts for 13a and 13c were referenced to tetramethylsilane computed at the same level of theory. The

13

C chemical shifts of 14a and 14c were calculated following similar

procedures.

Cytotoxic Activity Evaluation. The overall growth inhibitory effect of



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 44

terpecurcumins J–W (1–14) was evaluated against human breast cancer cells (MDA-MB-231 and MCF-7), human liver cancer cells (HepG2) and human lung cancer cells (A549) using crystal violet staining method as previously described.35 For cell cycle analysis, MCF-7 cells were treated with terpecurcumins LP (3–7) at 25 μM or terpecurcumin Q (8) at 5 and 7.5 μM for 48 h, and were then harvested for sub-G1 assay. Cells were washed two times with PBS and fixed with 75% ethanol at 20 C for 24 h. The cells were resuspended in PBS containing RNaseA (1 mg/mL), incubated for 1 h at 37 C, and then stained with propidium iodide (PI) (10 μg/mL) for 30 min at room temperature in the dark. The DNA contents of the stained cells were analyzed using ModFit LT software with FACSCalibur flow cytometry, and sub-G1 content stands for populations of cell apoptosis. For western-blotting analysis of mitochondrial-associated proteins

and

cleaved

PARP,

the

cell

lysate

was

prepared

in

ice-cold

radioimmunoprecipitation assay (RIPA) buffer. Cell lysates were resolved by electrophoresis and transferred to a PVDF membrane. The blot was then probed with primary

antibody

followed

by

incubation

with

appropriate

horseradish

peroxidase-conjugated secondary antibodies. The signal was visualized by enhanced chemiluminescence and recorded on an X-ray film.

LC/MS/MS Quantitative Analysis. Crude materials of Curcuma longa (14 commercial batches collected from different areas in China, including No. JH200912) was powdered and sifted (80-mesh). An aliquot of 2.0 g was mixed with 10 mL of methanol. The mixture was vortexed (2200 rpm) for 1 min, and then ultrasonicated (150 W, 40 kHz) in a water bath at 30°C for 60 min. The supernatant was passed through a 0.22-μm filter, and an aliquot of 10 μL was injected for LC/MS/MS analysis. Compounds 1, 3, 7, 8 and 13 were separated on a reversed-phase HPLC column, eluted with a gradient of acetonitrile/water/formic acid (Fig. S41), and then detected using a triple quadrupole mass spectrometer in (-)-SRM mode. The working solutions were prepared


Page 41 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


by serially diluting a stock solution of mixed standards (0.006, 0.018, 0.06, 0.18, 0.6, 1.8, 6 and 18 μg/mL for each compound). Regression equations were established by plotting the peak areas (duplicate injections) against nominal concentrations (Table S2).

ASSOCIATED CONTENT Supporting Information. Key HMBC and

1

H1H COSY correlations, NOE

enhancements, 1D and 2D NMR, IR and HRESIMS spectra for compounds 1−14; crystal cell diagram and CIF file for 1; and cytotoxicity evaluation and Western blotting analysis data. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *Tel./Fax: +86 10 82802024. E-mail: [email protected] (M. Ye). Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work was supported by National Natural Science Foundation of China (No. 81222054), the Program for New Century Excellent Talents in University from Chinese Ministry of Education (No. NCET-11-0019), and China State Administration of TCM (No. 201307002). We wish to thank Prof. Laurence A. Nafie (Department of Chemistry, Syracuse University/BioTools, Inc., USA) for the VCD measurement and calculation, and Dr. N. D. Wang (College of Chemistry and Molecular Engineering, Peking University) for his kind help with crystal structure analysis.

REFERENCES



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(1) Jemal, A.; Bray, F.; Center, M. M.; Ferlay, J.; Ward, E.; Forman, D. CA Cancer J. Clin. 2011, 61, 6990. (2) Printz, C. Cancer 2012, 118, 289291. (3) Chabner, B.A. N. Engl. J. Med. 2011, 365, 21472149. (4) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2012, 75, 311335. (5) Li, R.; Xiang, C.; Zhang, X.; Guo, D. A.; Ye, M. Curr. Pharm. Anal. 2010, 6, 256268. (6) Li, S.; Yuan, W.; Deng, G.; Wang, P.; Yang, P.; Aggarwal, B. B. Pharm. Crops 2011, 2, 2854. (7) Esatbeyoglu, T.; Huebbe, P.; Ernst, I. M.; Chin, D.; Wagner, A. E.; Rimbach, G. Angew. Chem. Int. Ed. 2012, 51, 53085332. (8) Gupta, S. C.; Prasad, S.; Kim, J. H.; Patchva, S.; Webb, L. J.; Priyadarsinic, I. K.; Aggarwal, B. B. Nat. Prod. Rep. 2011, 28, 19371955. (9) Teiten, M. H.; Gaigneaux, A.; Chateauvieux, S.; Billing, A. M.; Planchon, S.; Fack, F.; Renaut, J.; Mack, F.; Muller, C. P.; Dicato, M.; Diederich, M. OMICS 2012, 16, 289300. (10) Lin, X. H.; Ji, S.; Li, R.; Dong, Y. H.; Qiao, X.; Hu, H. B.; Yang, W. Z.; Guo, D. A.; Tu, P. F. Ye, M. J. Nat. Prod. 2012, 75, 21212131. (11) Xiao, Y. C.; Xie, J.; Yu, M.; Liu, M.; Ran, J.; Xi, Z.; Li, W.; Huang, J. Chin. Chem. Lett. 2011, 22, 1457–1460. (12) Xiao Y. C.; Lei J.; Liu M.; Yu, M.; Ran, J.; Xie, J.; Li, W.; Huang, J. Helv. Chim. Acta 2012, 95, 327332. (13) Li, W.; Wang, S. S.; Feng, J. T.; Xiao, Y. S.; Xue, X. Y.; Zhang, H.; Wang, Y. Q.; Liang, X. M. Magn. Reson. Chem. 2009, 47, 902–908. (14) Sy, L. K.; Brown, G. D. Magn. Reson. Chem. 1997, 35, 424425. (15) Koo, H. J.; Gang, D. R. PloS One 2012, 7, e51481. (16) Tuvi-Arad, I.; Avnir, D. Chem. Eur. J. 2012, 18, 1001410020. (17) Wang, J.; Chen, Z.Y.; Yan, B.C.; Ji, R.Y. Chin. J. Med. Chem. 1990, 1, 4953. (18) Mukherji, S. M.; Bhattacharyya, N. K. J. Am. Chem. Soc. 1953, 75, 4698–4700. (19) Qin, N. Y.; Yang, F. Q.; Wang, Y. T.; Li, S. P. J. Pharm. Biomed. Anal. 2007, 43, 486492.


Page 42 of 44

Page 43 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(20) Kreiser, W.; Körner, F. Helv. Chim. Acta 1999, 82, 16101628. (21) Golding, B. T.; Pombo-Villar, E. J. Chem. Soc., Perkin Trans 1 1992, 15191524. (22) Debie, E.; De Gussem, E.; Dukor, R. K.; Herrebout, W.; Nafie, L. A.; Bultinck, P. ChemPhysChem 2011, 12, 15421549. (23) Xie J. X. In The application of infrared spectroscopy in organic chemistry and medicinal chemistry, Science Press: Beijing, 1987, p. 331. (24) Qu, Y.; Xu, F. M.; Gao, H. Y.; Zhi, J. C. Y.; Wu, L. J. J. Shenyang Pharm. Univ. 2009, 26, 523526. (25) Carman, R. M.; Rayner, A. C. Aust. J. Chem. 1994, 47, 195202. (26) Liu, D. Z.; Wang, F.; Liao, T. G.; Tang, J. G.; Steglich, W.; Zhu, H. J.; Liu, J. K. Org. Lett. 2006, 8, 5749−5752. (27) Bifulco, G.; Gomez-Paloma, L.; Riccio, R. Tetrahedron Lett. 2003, 44, 7137–7141. (28) Chini, M. G.; Jones, C. R.; Zampella, A.; DAuria, M. V.; Renga, B.; Fiorucci, S.; Butts, C. P.; Bifulco, G. J. Org. Chem. 2012, 77, 1489−1496. (29) Rho, J. R.; Hwang, B. S.; Sim, C. J.; Joung, S.; Lee, H. Y.; Kim, H. J. Org. Lett. 2009, 11, 55905593. (30) Forestieri, R.; Merchant, C. E.; de Voogd, N. J.; Matainaho, T.; Kieffer, T. J.; Andersen, R. J. Org. Lett. 2009, 11, 51665169. (31) Huang, J. H.; Yang, J. R.; Zhang, J.; Yang, J. J. Am. Chem. Soc. 2012, 134, 8806−8809. (32) Liu, L.; Bruhn, T.; Guo, L.; Götz, D. C.; Brun, R.; Stich, A.; Che, Y.; Bringmann, G. Chem. Eur. J. 2011, 17, 26042613. (33) Rocha, D. F. O.; Hamilton, K.; Gonçalves, C. C. S.; Machado, G.; Marsaioli, A. J. J. Nat. Prod. 2011, 74, 658–663. (34) Kim, H. J.; Ruszczycky, M. W.; Choi, S. H.; Liu, Y. N.; Liu, H. W. Nature 2011, 473, 109–112. (35) Zhao, C.; Yin, S. T.; Dong, Y. H.; Guo, X.; Fan, L. H.; Ye, M.; Hu, H. B. Autophagy 2013, 9, 196207. (36) Matthes, H. W. D.; Luu, B.; Ourisson, G. Phytochemistry 1980, 19, 2643–2650.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(37) Sybyl Software, Version X 1.1, Tripos Associates Inc., St. Louis, MO, 2010. (38) Bruhn, T.; Hemberger, Y.; Schaumlöffel, A.; Bringmann, G. SpecDis, Version 1.51, University of Wuerzburg, Germany, 2011. (39) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, Jr. J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli,C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision B.01, Gaussian Inc., Pittsburgh, PA, 2010. (40) Debie, E.; Bultinck, P.; Nafie, L. A.; Dukor, R. K. CompareVOA, BioTools Inc., Jupiter, FL, 2010.


Page 44 of 44

Density Functional Theory Calculations in Stereochemical

Recommend Documents